The main aim of this project is to develop an appropriate and accurate dietary assessment technique for the African American population that will enhance our understanding of the role of diet in the etiology of cancer and the role of diet in the differential cancer rates of African Americans and Caucasian Americans. In this study we propose to: a) develop and cognitively test a food frequency questionnaire (FFQ) to measure dietary intake among various SES levels of the African American population, using the dietary data collected through 24-hour recalls; and b) validate/calibrate the new tool using the same population. We will develop an FFQ utilizing 1,200 24-hour recalls, which will be collected from hospital employees and public housing residents. The questionnaire will be cognitively tested utilizing "think aloud" interviewing method, in which, respondents verbalize their thought process while answering several variations of the questionnaire. To validate the newly developed instrument, 336 new subjects will be selected: 168 from public housing communities, and the other 168 will be employees of the hospitals. Initially we will administer the new FFQ to all subjects. Next, three 24- hour recalls will be collected from each subject, and compared with the results of the FFQ. Finally, comparison with the commonly used Block and Willett FFQs will be made.